The present invention relates to a method of communicating across a telecommunications network and associated apparatus. In particular, the invention relates to a method of communicating across a telecommunications network, to a telecommunications network, to a network management system for setting connections in a network and to a network element of such networks.
Telecommunications networks, particularly optical networks, have in the past routed data across the network by setting routes by means of a manually operated network management system. When changes are required to be made to the route or routes set in the network, response times can be very long in comparison to the rate of transmission of data.
Significant improvements in routing of data have been made in recent years in the context of electrical networks. One such improvement is the ability of network elements of the network to route data packets without reverting to a separate network management system. Recently, the use of Multi Protocol Label Switching (MPLS), which is currently used in both IP and ATM networks, has been recognised as being particularly advantageous. One of the benefits of MPLS is that network elements of the network are able to route a given data packet quickly, by reference to a label in the data packet. Furthermore, since the routing of data packets does not require the exchange of data with a network management system the use of MPIS has a major advantage in that it facilitates dynamic network control without the delays often associated with networks controlled by a network management system.
It has been proposed, so as to facilitate dynamic network control, to introduce MPLS, in the form of a Generalised Multi-protocol Label Switching (GMPLS) method, into optical networks. However, incorporating GMPLS into an optical network is not straightforward. Two proposals have been made for implementing GMPLS in an optical network as will now be described.
The first proposal may be referred to as the “Peer-to-Peer Model” and is illustrated by FIG. 1 of the accompanying drawings. With reference to FIG. 1, a first IP network 1 is connected via an optical network 2 to a second IP network 3. The optical network 2 is required to make available to the IP networks 1, 3 topological information (in the form of IP information) so that data packets can be routed from the first IP network 1 to the second IP network 3 via the optical network 2 by means of IP data in the data packet. If the optical network 2 is privately owned, making such topological information publicly available may however be undesirable. For example, such information may be considered to be commercially sensitive and it may be desired to keep such information confidential.
The second proposal, which may be referred to as the “Client-Server Model”, does not require the optical network to make public such topological information. In this second proposal, which is described with reference to FIG. 2 of the accompanying drawings, a first IP network 1 is connected via an optical network 2 to a second IP network 3 in a manner similar to that of the first proposal. However, in this proposal the interfaces between the first and second IP networks 1, 3 (the clients) and the optical network 2 (the server) each include a user network interface 4 (UNI). Thus the first IP network 1, via a first UNI 4a, effectively request a connection across the optical network 2 by means of IP data in a data packet. Topology information relating to the optical network 2 is however not made available outside the optical network 2.
Both of the proposals described above suffer from a significant disadvantage. In order for the optical network to operate in a GMPLS environment it is necessary, in the proposals made, for the network elements of the optical network to process and handle network topology information and to set up and tear down network connections. In order for the individual network elements to be able to perform such tasks the network elements each require significant processing capability and access to significant amounts of memory. Whilst such requirements can be met when installing new optical networks, many existing optical network elements are not able to perform at the required level. Replacing such existing optical networks (often referred to as legacy networks) would be costly and is therefore undesirable.